mirror of
https://github.com/DSpace/DSpace.git
synced 2025-10-23 09:53:09 +00:00
c13f24e570
dspace/etc to overwrite the originals for the Oracle port to work. The must be kept up to date with any schema changes! git-svn-id: http://scm.dspace.org/svn/repo/trunk@1060 9c30dcfa-912a-0410-8fc2-9e0234be79fd
108 lines
4.4 KiB
Plaintext
108 lines
4.4 KiB
Plaintext
DSpace on Oracle
|
|
Revision: 11-sep-04 dstuve
|
|
|
|
|
|
1 Introduction
|
|
2 Installing DSpace on Oracle
|
|
3 Porting notes for the curious
|
|
|
|
|
|
DSpace is now tested to work with Oracle 10x, and should work with
|
|
version 9 as well. The installation process is quite simple, involving
|
|
changinge a few configuration settings in dspace.cfg, and replacing a few
|
|
Postgres-specific sql files in etc/ with Oracle-specific ones. Everything
|
|
is fully functional, although Oracle limits you to 4k of text in text fields
|
|
such as item metadata or collection descriptions.
|
|
For people interested in switching from Postgres to Oracle, I know of
|
|
no tools that would do this automatically. You will need to recreate the
|
|
community, collection, and eperson structure in the Oracle system, and then
|
|
use the item export and import tools to move your content over.
|
|
|
|
How to Install DSpace on Oracle
|
|
|
|
Things you need
|
|
|
|
Oracle 9 or 10
|
|
Oracle JDBC driver
|
|
sequence update script from Akadia consulting:
|
|
currently incseq.sql at: http://akadia.com/services/scripts/incseq.sql
|
|
(put in your dspace source /etc directory)
|
|
|
|
Installation procedure
|
|
|
|
1. Create adatabase for DSpace. Make sure that the character set is one
|
|
of the Unicode character sets. DSpace uses UTF8 natively, and I would
|
|
suggest that the Oracle database use the same character set. Create
|
|
a user account for DSpace (we use 'dspace',) and ensure that it has
|
|
permissions to add and remove tables in the database.
|
|
|
|
2. Edit the config/dspace.cfg file in your source directory for the
|
|
following settings:
|
|
|
|
db.name = oracle
|
|
db.url = jdbc.oracle.thin:@//host:port/dspace
|
|
db.driver = oracle.jdbc.OracleDriver
|
|
|
|
3. Go to etc/oracle in your source directory and copy the contents
|
|
to their parent directory, overwriting the versions in the parent:
|
|
|
|
cd dspace_source/etc/oracle
|
|
cp * ..
|
|
|
|
You now have Oracle-specific .sql files in your etc directory, and
|
|
your dspace.cfg is modified to point to your Oracle database and
|
|
are ready to continue with a normal DSpace install, skipping the
|
|
Postgres setup steps.
|
|
|
|
NOTE**** DSpace uses sequences to generate unique object IDs - beware
|
|
Oracle sequences, which are said to lose their values when doing
|
|
a database export/import, say restoring from a backup. Be sure
|
|
to run the script etc/udpate-sequences.sql.
|
|
|
|
|
|
Oracle Porting Notes for the Curious
|
|
|
|
Oracle is missing quite a number of cool features found in Postgres, so
|
|
workarounds had to be found, most of which are hidden behind tests of
|
|
the db.name configuration parameter in dspace.cfg. If the db.name is
|
|
set to Oracle the workarounds are activated:
|
|
|
|
Oracle doesn't like ';' characters in JDBC SQL - they have all been removed
|
|
from the DSpace source, including code in the .sql file reader to strip ;'s.
|
|
|
|
browse code - LIMIT and OFFSET is used to limit browse results, and an
|
|
Oracle-hack is used to limit the result set to a given size
|
|
|
|
Oracle has no boolean data type, so a new schema file was created that
|
|
uses INTEGERs and code is inserted everywhere to use 0 for false
|
|
and 1 for true if the db.name is Oracle
|
|
|
|
Oracle doesn't have a TEXT data type either, so TEXT columns are defined
|
|
as VARCHAR2 in the Oracle-specific schema.
|
|
|
|
Oracle doesn't allow dynamic naming for objects, so our cute trick to
|
|
derive the name of the sequence by appending _seq to the table name
|
|
in a function doesn't work in Oracle - workaround is to insert Oracle
|
|
code to generate the name of the sequence and then place that into
|
|
our SQL calls to generate a new ID.
|
|
|
|
Oracle doesn't let you directly set the value of sequences, so
|
|
update-sequences.sql is forced to use a special script sequpdate.sql
|
|
to update the sequences.
|
|
|
|
Bitstream had a column 'size' which is a reserved word in Oracle,
|
|
so this had to be changed to 'size_bytes' with corresponding code changes.
|
|
|
|
VARCHAR2 has a limit of 4000 characters, so DSpace text data is limited to 4k.
|
|
Going to the CLOB data type can get around that, but seemed like too much effort
|
|
for now. Note that with UTF-8 encoding that 4k could translate to 1300
|
|
characters worst-case (every character taking up 3 bytes is the worst case
|
|
scenario.)
|
|
|
|
DatabaseManager had to have some of the type checking changed, because Oracle's
|
|
JDBC driver is reporting INTEGERS as type DECIMAL.
|
|
|
|
Oracle doesn't like it when you reference table names in lower case when
|
|
getting JDBC metadata for the tables, so they are converted in TableRow
|
|
to upper case.
|